Effective optical resolution of Epson 4990

[Rider]

I know the Epson 4990 is popular with people here. I was wondering whether you recall what it's effective optical resolution is. Thanks.

[sanking]

I know the Epson 4990 is popular with people here. I was wondering whether you recall what it's effective optical resolution is. Thanks.

1600 - 1800 dpi.

Sandy

[Rider]

Thanks Sandy.

Is the 4990 color sensitive enough that it would be worthwhile getting IT8 targets for it?

[Rider]

Wow, the 4990 Pro included iT8 targets. http://www.epson.com/cmc_upload/0/000/054/482/P4990_InfoSheet.pdf

Does that answer my question?

[Brian Ellis]

Tests by the late Ted Harris published in View Camera magazine showed 2100-2200 ppi. I've seen the same number in other reviews and in posts by others here.

[BetterSense]

The limited tests that I ran showed that you don't gain much by scanning over 2400dpi. For what it's worth.

[domaz]

The limited tests that I ran showed that you don't gain much by scanning over 2400dpi. For what it's worth.

I think that's true but I have found you get slightly better quality if you scan at the maximum dpi and then downsample. I have Vuescan set to scan at 4800 dpi and then downsample to 2400 dpi. Setting the scanner to 4800 dpi seems to "tell" the scanner to use the maximum quality available for the scan.

[Rider]

Ok, assuming the number is between 1600 and 2200, how would you squeeze the most of this scanner--scan at 2400, or scan at 4800 and downsample to 2400 using scanning software (i.e. output at 2400)?

[sanking]

Ok, assuming the number is between 1600 and 2200, how would you squeeze the most of this scanner--scan at 2400, or scan at 4800 and downsample to 2400 using scanning software (i.e. output at 2400)?

First, shim your film holder to make sure that the film is at the optimum plane of focus.

Second, scan at the highest possible resolution, 4800, then downsize to 2400.

Sandy King

[Rider]

Sandy,

Can you explain what "shim the film holder" means please? Thanks!

[BetterSense]

First, shim your film holder to make sure that the film is at the optimum plane of focus, and/or buy bettescanning.com holders.

Second, use glass if necessary to ensure the film is flat, especially if it has any curl.

Third, scan at the highest possible resolution, 4800, then downsize to 2400, assuming that you can afford to wait that long (it's very slow at 4800).

[Rider]

First, shim your film holder to make sure that the film is at the optimum plane of focus, and/or buy bettescanning.com holders.

Second, use glass if necessary to ensure the film is flat, especially if it has any curl.

Third, ...

I was looking at betterscanning, but doesn't their LF holder require that you use glass even when dry mounting?

[Ivan J. Eberle]

The Epson 4x5 holder is pretty good. At least for the one 4990 I've used, it was set at the optimum height for best focus straight from the factory.

Don't have the 4990 to use anymore, but I seem to recall hearing that there needs to be a film holder in place to achieve the highest resolution setting of the internal optics--that the holder has a tab that sets a switch (else the resolution max is 1800 dpi, as when scanning off the glass sans holder)?

[sanking]

Sandy,

Can you explain what "shim the film holder" means please? Thanks!


By shim I just meant use washers, coins or screws, whichever applies to your holder, to place the negative at the right focus point.

Sandy

[Ivan J. Eberle]

It's really easy to scratch a scanner's platen glass with washers coins and screws... with the lid pressing down on them. Nylon washers and screws are a safer bet than metal ones (or coins of any type). And, once again, I didn't need them with the 4990 with the stock holders.

[Nathan Potter]

Yes to Ivan. I use pieces of thin mat board cut into 1/4 inch squares to raise the film plane off the glass platen. Protect that platen from scratches at all costs. I found with a V750 that there is considerable depth of field, 1.5 mm or so, about the plane of best focus within which you'll find 2000 to 2400 dpi resolution. This depth of field may be even greater with the OPs' machine.

Nate Potter, Austin TX.

[Rider]

How do you determine the best focus? In the darkroom, it was very easy to focus on the grain.

[sanking]

How do you determine the best focus? In the darkroom, it was very easy to focus on the grain.

Use a very sharp negative with lots of detail, or a resolution target if you have one. Scan the negative at the highest dpi, with the negative located at varying distances from the glass, say on the glass, .5mm above the glass, 1mm above the glass, 1.5mm above the glass, 2.5mm above the glass, etc. Examine each scan and compare to determine which distance gives highest sharpness.

Depth of field is very great with the lens of the 4990, also with the lens of the V700/750 that is enabled when you choose film area guide. DOF is much less with the lens of the V700/V750 that is enable when you choose film holder.

Sandy

[W K Longcor]

If I may barge into this thread with some possibly related questions --- I have some antique glass lantern slides that I want to scan on a newly aquired 4990. Does the focus change with the setting of either film in frame or on glass with the film area guide? If so, am I better setting the software for use of the frame -- even hough the slide is just on the glass platten -- since the thickness of the glass slide mount raises the image a bit? ( I hope that jumbled sentence makes sence to someone).
Also, is there any difference between the 4990 photo and the 4990 pro other than supplied software?

[Nathan Potter]

How do you determine the best focus? In the darkroom, it was very easy to focus on the grain.

Yes, as Sandy says. It is of course time consuming depending on how many distances above the platen you want to examine. And that depends upon how fussy you might want to be.

Well here's a bit of technique as practiced in the IC industry:
The classic way to determine plane of best focus is to use a glass wedge resolution mask, which must be purchased and is expensive. Short of that you could make one by exposing an array of fine lines on fine grained film. The film negative is then taped to a blank piece of glass and made as flat as possible. It can be 2X2 or 4X4 inch, or whatever size is convenient. Now this is placed on the scanner platen with the film emulsion side down on the bottom of the glass. But instead of being flat against the scanner platen one end of the film plate will rest on shims while the other end rests directly on the scanner platen. Thus the slope of the film will run through the plane of best focus somewhere along the slightly angled film. I'd have the shimmed end at least 6 mm above the platen with the other end resting on the platen. Do a scan and observe the image on the monitor. You will easily see the sharpest lines some where along the image and you can easily calculate how high above the platen that part of the image was as long as your film was reasonably tight against the glass. Plus you should get some idea of the depth of field in the scanner imaging system.

Ordinarily the technique is done using a glass resolution mask with repeating test target cells arrayed over the surface and with known line width dimensions so that actual resolution numbers can be extracted. But such all glass plates are very expensive.

I suppose you could make things even simpler by just placing scratches directly on a piece of glass using a diamond scribe or perhaps trying to image a piece of ground glass - the advantage being to avoid possible film sag. But if your film won't sit flat enough use some thin double stick tape under it in selected places.

Nate Potter, Austin TX.

[David Low]

First, shim your film holder to make sure that the film is at the optimum plane of focus, and/or buy bettescanning.com holders.

Second, use glass if necessary to ensure the film is flat, especially if it has any curl.

Third, scan at the highest possible resolution, 4800, then downsize to 2400, assuming that you can afford to wait that long (it's very slow at 4800).

Apologies in advance for the dumb question, but if you scan at 4800, how do you then downsize the image to 2400?

[Rider]

Apologies in advance for the dumb question, but if you scan at 4800, how do you then downsize the image to 2400?

The best way to do it is in the scanning software.

Vuescan has an output size reduction feature.

Under input, you select 4800 dpi.

Then under output you select "TIFF size reduction factor" of 2 (assuming you're scanning to TIFF).

[Rider]

Use a very sharp negative with lots of detail, or a resolution target if you have one. Scan the negative at the highest dpi, with the negative located at varying distances from the glass, say on the glass, .5mm above the glass, 1mm above the glass, 1.5mm above the glass, 2.5mm above the glass, etc. Examine each scan and compare to determine which distance gives highest sharpness.

Depth of field is very great with the lens of the 4990, also with the lens of the V700/750 that is enabled when you choose film area guide. DOF is much less with the lens of the V700/V750 that is enable when you choose film holder.

Sandy

Thank you. As a reality check, what range of mm do people see for the 4990?

[sanking]

Thank you. As a reality check, what range of mm do people see for the 4990?

I can not speak to other "people" but I have tested several different 4990 scanners and got a maximum of 1600-1800 dpi effective resolution. That is 60-70 l/pm (or 31-35 lp/mm).


Sandy King

[Rider]

Sandy--my apologies if it wasn't clear. I was referring to mnm of shimming.

[sanking]

Sandy--my apologies if it wasn't clear. I was referring to mnm of shimming.

Sorry about the misunderstanding. I think you will find that best plane of focus with the 4990 is between 0.5mm to 2.0 mm above the top of the scanner glass.

Sandy